University of Southampton
Modelling and managing the charging of massed electric vehicles on constrained residential power networks
Abstract
dc:description.abstractThis thesis investigates how Electric Vehicles (EVs) and UK residential 230 V Low Voltage (LV) networks interact, as EVs supplant fossil-fuel cars. Home EV charging can provoke LV overloads, invisible to analyses set at higher voltages using averaged data. The Feeder Phase Balancer (FPB) simulator is presented, which models EVs on an LV network. FPB simulates various EVs driving independent trips, with charging control by a local manager and/or a 3rd party Aggregation service. Vehicle to Grid, Demand Response, Fast Frequency Response and Time of Use services are modelled. Impacts on EV and LV conditions are recorded; the EV support capability of various LV networks is explored and remedial options assessed. New knowledge is found concerning seasonal EV need to charge and of ability of networks to support Winter / Summer charging. Recommendations to DNOs and interested parties are formed. The FPB has simulated over 4.9 x 10⁹ independent EV trips, finding: a) LV networks, built to historic guidelines for gas-heated homes, cannot support EVs replacing cars 1:1 without reinforcement and / or local ICT control; b) Replacing gas-heating with Heat Pumps adds far greater load, forcing network reinforcement for the great majority of UK LV networks. c) Localised LV overloads (faults and blackouts) rise with EV charging duties; d) Network conditions, EV numbers and characteristics, ambient temperature and driven distances affect overloads, seen as EVs reach 10% ~ 30% homes e) There is seasonality: cold Winters provoke overloads (possibly simultaneous) f) Local ICT management of charging helps limit overloads, but can cause undercharging. Vehicle to Grid EVs can reduce undercharging. g) Top-down control is found not beneficial e.g. Time of Use constraints can cause undercharging when imposed and “crowd-rush” overloads when lifted; h) Destination charger use has great seasonality (x4 to x22 rise in Winter); failing to charge threatens stranding vehicles in the evenings. The above outcomes are dependant upon assumptions (e.g. battery characteristics) which are described. A repository of results data is available.
Degree
thesis:*- Name dc:type.qualificationname
- Ph.D.
- Level dc:type.qualificationlevel
- doctoral
- Year dc:date.issued
- 2020
Author and committee
dc:creator, dc:contributor.*- Author dc:creator
-
- Broderick, Stephen Redmond
- Advisor dc:contributor.advisor
-
- Chipperfield, Andrew
- ORCID person:identifier.orcid
Rights
- Language dc:language
- English, eng